RMA - Week 3

Experiment

• manipulation fo one or more variables

• determining effect of manipulation

• test cause-effect of both variables (causality)

Nuisance variables

• An addictional factor that affects dependent variable

• Need to be considered when setting up methodology

Solution - nuisance variables

• turn into control variable - make same for everyone

• prove that they dont affect DV

• variation across levels of indpeendent - then hold variable constant for all pps

• across pps - randomly assign pps

• counterbalancing

• include control

Coufounding vs nuisance

• If nusiance variavle varies across xonditions it becomes cofounding

experiemental group

group recieving the most importnanr level of the independent variable

2+ IVs

• Can have multiple IV's

• If we cant control something - acknowledge it as another IV

• Can get rid of nuisance variables

• More efficient

• More accurate results

2+ DVs

• Dependent variables (DVs) are measures of the behaviour in question.

• Measuring more than one DV is usually more informative.

S/W experiments

• S

• Relative strong test of causality

• Possibility of a variety of manipulative controls.

 

W

•  Unnatural setting and tasks.

•  Reactivity (also in non-experimental research).

• Some phenomena cannot be studied under controlled conditions

• Ethical limitations